The general goal of the proposed work is to test a theory that links the COMT and GABRA2 genes to intermediate phenotypes, and, in turn, to the important clinical problem of relapse to substance abuse. It will test whether genes that have been empirically linked to substance dependence, and to measures of frontal brain function (viz., fast beta power in the spontaneous electroencephalogram and frontal P300a amplitude), also confer an increased risk for relapse to these disorders. The specific goals of the project are: (1) to examine whether the genotypes of 100 cocaine-, heroin, or polydrug-dependent patients who return to substance use within 4 months after study enrollment are different from those of 100 patients who successfully maintain abstinence and 50 non-substance-dependent controls;(2) to replicate our previous findings of enhanced electroencephalographic fast beta activity and reduced frontal P300a amplitude in patients who return to substance use in comparison to patients who maintain abstinence and to healthy non-substance-dependent controls;(3) to determine if polymorphisms in GABRA2 and COMT genes are respectively associated with phenotypic variation in EEG fast beta power and frontal P300a amplitude;(4) to determine if genetic markers improve the prediction of relapse beyond the predictive accuracy attained with EEG fast beta power and frontal P300a amplitude, in combination with other known risk factors, including severity/chronicity of dependence, age, type of substance dependence, and Antisocial Personality Disorder.